Table 2 ROC analysis of potential biomarkers with the highest AUC scores from analysed datasets.

From: Characterization of different fat depots in NAFLD using inflammation-associated proteome, lipidome and metabolome

Predictors

AUC

95% confidence interval

SN

SP

Threshold

Lower limit

Upper limit

Myocardial TG

Waist

0.905

0.810

1.000

0.92

0.84

90.5 cm

CDCP1

0.837

0.725

0.949

0.80

0.76

3.31

CE 20:3

0.858

0.748

0.967

0.88

0.80

0.804

Glucose

0.845

0.734

0.955

0.76

0.84

3.5 × 108

Epicardial FAT

Visceral fat

0.952

0.899

1.000

0.92

0.88

1373 mm2

IL - 18R1

0.853

0.748

0.957

0.84

0.80

6.69

CE 20:3

0.888

0.793

0.983

0.92

0.76

0.832

*PE (18:0/20:4)

0.850

0.734

0.965

0.80

0.88

6.5 × 105

Pericardial FAT

Waist

0.975

0.927

1.000

1.00

0.92

93 cm

HGF

0.963

0.919

1.000

0.96

0.88

6.79

diSM 18:0

0.938

0.864

1.000

0.92

0.88

0.225

*CER (d18:0/16:0)

0.942

0.869

1.000

1.00

0.8

8.7 × 105

Liver FAT

Visceral fat

0.984

0.960

1.000

0.92

0.96

1588.5 mm2

CDCP1

0.926

0.844

1.000

0.84

0.96

3.50

diSM 18:0

0.963

0.917

1.000

0.92

0.88

0.226

Glutamate

0.946

0.874

1.000

0.96

0.92

2.9 × 108

  1. Unless otherwise stated, all threshold values represent relative units. *PE (18:0/20:4) – 1-stearoyl-2-arachidonoyl-GPE (18:0/20:4); *CER (d18:0/16:0) – N-palmitoyl-sphinganine (d18:0/16:0).